The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses dynamic context injection (! ls -la docs/design/*.md | head -10) to retrieve a list of design documents when the skill is loaded. This is a legitimate use of shell execution for project context discovery and does not involve sensitive directories or network access.
[SAFE]: Analysis of the execution process confirms that the skill follows a structured orchestration pattern, relying on defined sub-agents for specialized tasks like test generation and planning, with no attempts to bypass safety filters or exfiltrate data.

recipe-plan